DNA amplification is a process of replicating a target double-stranded DNA (dsDNA) to generate multiple copies of it. Since individual strands of a dsDNA are antiparallel and complementary, each strand may serve as a template strand for the production of its complementary strand. The template strand is preserved as a whole or as a truncated portion and the complementary strand is assembled from deoxynucleoside triphosphates (dNTPs) by a DNA polymerase. The complementary strand synthesis proceeds in 5′→3′ direction starting from the 3′ terminal end of a primer sequence that is hybridized to the template strand.
Whole-genome amplification (WGA) involves non-specific amplification of a target DNA. WGA is often achieved by multiple displacement amplification (MDA) techniques employing random oligonucleotide primers for priming the DNA synthesis at multiple locations of the target DNA along with a high fidelity DNA polymerase having a strand displacing activity (e.g., Phi29 polymerase). Even though currently available commercial WGA systems such as GenomiPhi (GE Healthcare, USA) and RepliG (Qiagen) kits provide optimal results with high molecular weight target DNA, performance of these systems is poor when the target DNA is short and/or highly fragmented. When the target DNA is fragmented and the sequence length is less than about 1000 nucleotides, amplification of the target DNA using conventional methods results in decreased amplification speed, significant sequence dropout especially near the ends of the target DNA, and highly sequence-biased amplification. As the length of the template DNA is decreased, the likelihood of that strand being primed multiple times decreases in the MDA reaction. This decreases the amplification potential of these shorter fragments. Efficient methods for non-specifically amplifying short, fragmented DNA are therefore highly desirable.
Ligation-mediated polymerase chain reaction (PCR) has been used to amplify fragmented dsDNA. However, only a small fraction of the fragmented DNA gets amplified in these reactions leading to inadequate genome coverage. To efficiently amplify fragmented, target dsDNA, they may first be repaired and then be concatamerized by blunt-end ligation to generate sequences that are longer than 1000 base pairs (bp). However, a relatively higher concentration of the target DNA is often required to promote concatamerization and subsequent amplification. Circularization of double-stranded target DNA has also been employed in various nucleic acid based assays including MDA, WGA, hyper-branched rolling circle amplification (RCA) and massively parallel DNA sequencing. To effectively circularize and amplify fragmented dsDNA, the double-stranded ends of the fragmented DNA are first repaired, followed by blunt-end ligation to form double-stranded DNA circles. However, it is difficult to circularize double-stranded DNA fragments that are less than 500 bp in length.
The double-stranded DNA may be denatured to produce single-stranded DNA (ssDNA), which may further be circularized in a template-dependent intra-molecular ligation reaction using a ligase. However, prior sequence information of the target DNA is required to perform a template-dependent circularization. Template-independent intra-molecular ligation of ssDNA has also been documented. For example, TS2126 RNA ligase (commercially available under the trademarks THERMOPHAGE™ RNA ligase II or THERMOPHAGE™ ssDNA ligase (Prokaria, Matis, Iceland) or CIRCLIGASE™ ssDNA ligase (Epicenter Biotechnologies, Wisconsin, USA) has been used for making digital DNA balls, and/or locus-specific cleavage and amplification of DNA, such as genomic DNA. CIRCLIGASE I™ has a low degree (about 30%) of adenylation where as CIRCLIGASE II™ comprises a substantially adenylated form of TS2126 RNA ligase. Linear, single-stranded complementary DNA (cDNA) molecules prepared from 5′-end fragments of mRNA have also been amplified via rolling circle replication after circularization using TS2126 RNA ligase. By appropriately incorporating a sense RNA polymerase promoter sequence in to the cDNA, the circularized cDNA template has shown to act as a transcription substrate and thus effect the amplification of the mRNA molecules in a biological sample. Further, the TS2126 RNA ligase has been used for amplifying the cDNA ends for random amplification of cDNA ends (RACE). From limited amounts of fragmented DNA, DNA template for rolling circle amplification has also been generated by employing TS2126 RNA ligase. The method involved denaturing the linear, fragmented dsDNA to obtain linear ssDNA fragments, ligating the linear ssDNA with CIRCLIGASE™ ssDNA ligase to obtain single-stranded DNA circle, and then amplifying the single-stranded DNA circle using random primers and Phi29 DNA polymerase via RCA. However, even after optimizing the reaction conditions, the amount of generated single-stranded circular DNA was highly variable and sequence dependent. For example, oligonucleotides comprising a 5′G and a 3′T nucleotide ligated significantly better than its complementary oligonucleotide comprising a 5′A and a 3′C under identical ligation conditions. Further, intra-molecular ligation efficiency varied among linear ssDNA sequences having identical or very similar sizes but with small differences in nucleotide sequence. The efficiency also varied among linear ssDNA sequences of different sizes (e.g., sequence length ranging from 100 bases to kilobases in size). Moreover, all attempts of ligation-amplification reactions involved intermediate isolation, purification and/or cleaning steps, thus making the ligation-amplification workflow cumbersome. For example, analysis of forensic samples of fragmented DNA by circularization followed by rolling circle amplification was carried out in multiple steps comprising 5′ DNA phosphorylation, adapter ligation, DNA circularization, and whole-genome amplification. Each step reactions were subjected to a reaction clean-up before performing the next step. Further, the multi-step process often resulted in the loss of template DNA and led to failed analysis. No amplification advantage was observed when ligation and amplification was performed in single reaction vessel; rather the components of the ligation reaction that were carried forward were often found to be inhibitory for the subsequent amplification reaction. Therefore, efficient methods for non-specifically amplifying short DNA sequences in a single reaction vessel without any intervening cleaning steps are highly desirable, especially in cases where representative and balanced whole genome information is desired. Further, methods for amplifying a linear nucleic acid sequence via rolling circle amplification in a single reaction vessel that overcome the inhibition caused by the reactants in each of the ligation-amplification are highly desirable.